ok so like.. this one was mostly for testing on my "ability" to do loras with a tiny dataset. i dont expect this one to get lots of attention, so just ignore this one. id put it private but of course i need to have a membership. also, i mixed in 2 artstyles so it can either be the the pixel one or the other one. now, i know that the images look good when theyre not zoomed in, but when you look closer then it looks a bit wonky but whatever thats ai for you i guess. dont blame me if it doesnt meet "your requirements" if the lora is bad, this wasnt meant to be public anyway but i need it on site for something.
UPDATE!!!! - apparently it works as a concept lora too if you lower the weights of this one very slightly and make the desired style a bit bigger on weights, examples are the last 3 images at the end of the showcase
now if you DO end up trying this one out for some reason, here is some tags:
kknocksstyle - activation tag
conjoined - kinda necessary
simple background and/or white background - kinda useful but not necesary i think
1girl - put that in every single image, do not add solo.
same conjoinment - basically same head like a clone
side conjoinment - basically can be different or same people with skin color different on side, for example dark on left and white on right
multiple conjoinment - basically different heads
Description
Details
Available On (1 platform)
Same model published on other platforms. May have additional downloads or version variants.